A novel method for retinal vessel segmentation and diameter measurement using high speed video

Area of Science:

• Ophthalmology
• Biomedical Engineering
• Medical Imaging

Background:

• Static retinal imaging provides limited insight into dynamic pathophysiological changes.
• Dynamic analysis of retinal vessels is crucial for understanding conditions like glaucoma and hypertension.
• Few methods exist for dynamic analysis of retinal video sequences.

Purpose of the Study:

• To develop and validate a novel method for assessing time-dependent diameter changes in rat retinal vessels.
• To enable dynamic analysis of high-speed retinal videos (125 fps).
• To provide tools for better understanding of retinal vascular abnormalities.

Main Methods:

• A novel method combining vessel segmentation and diameter measurement modules was developed.
• High-speed videos (125 fps) of rat retinas were utilized.
• Quantitative analysis of vessel segmentation (95.1% specificity, 97.3% sensitivity) and diameter measurement errors (-0.147±0.41 pixels) was performed.

Main Results:

• The proposed method accurately segments retinal vessels with high specificity and sensitivity.
• Diameter measurement errors were minimal, with a mean of -0.147±0.41 pixels.
• The study successfully demonstrated the assessment of dynamic, time-dependent diameter changes.

Conclusions:

• The novel method provides a reliable tool for dynamic retinal vessel analysis.
• This technique can enhance the understanding of pathophysiological changes in retinal vascular diseases.
• The findings support the use of dynamic retinal imaging in preclinical research for glaucoma and hypertension.